Geometry-Informed Data Representations
Workshop proceedings arising from the 1st-place Theory and Methods Challenge Fortnights award (£30,000). Explores geometry-informed representations of complex data, connecting …
Workshop proceedings arising from the 1st-place Theory and Methods Challenge Fortnights award (£30,000). Explores geometry-informed representations of complex data, connecting …
A comprehensive review and guide to network embedding methods, covering spectral, random walk, deep learning, and geometric approaches, with a unified framework for comparison.
📖 _Communications Physics_ · 2022 — A universal framework for multilayer network exploration using Random Walk with Restart, enabling the integration of heterogeneous biological …
Multilayer models and exploration methods for biological networks. Developed new formalism of random walk with restart on multilayer networks, the MultiXrank Python package, and …
Combined X-ray, neutron scattering and molecular simulation study revealing the microscopic origin of the pre-peak in aqueous propylamine mixtures.
We develop a theoretical framework for modelling micro-heterogeneity in liquid mixtures, examining the role of many-body correlations in determining mixture structure.